Application of AI techniques to IT Ops — Richard Whitehead // Moogsoft Inc.

Quotes

• "We tend to see artificial intelligence as software or computing replicating or at least trying to mimic human behavior. When you apply it to IT operations, it's a little less fancy." - Richard Whitehead

• "When we look at artificial intelligence for operations, it is the application of artificial intelligence techniques, programming techniques, like machine learning and natural language processing, to the business of the traditional industry, IT operations." - Richard Whitehead

• "One of the things we do as part of our product offering is offer something we call a situation. When an incident is detected, it's a workspace where individuals can collaborate and work on a problem." - Richard Whitehead

• "It's fraught with issues. So the first issue is one of intellectual property and privacy. You must be very cognizant that when you ask ChatGPT a question, you don't put anything proprietary or sensitive. Because that information is no longer yours, it's now in the public domain." - Richard Whitehead

• "When dealing with technologies that are very new and very advanced, perhaps developing code using very new programming languages, ChatGPT will struggle to provide information because it's not answering a question based on its understanding of technology. It uses a pre-trained model of other people's conversations." - Richard Whitehead

• "When we talk about applied machine learning and our techniques, the goal is always to make life easier for the human operator. In other words, can we shorten the cycle it takes for them to make a decision? Can we shorten the time it takes them to find the correct answer to the problem?" - Richard Whitehead

• "One thing that has distinguished ALOps from preceding technologies is that you can automate anything. We had a code available 20 or 30 years ago. Okay, CPU resources were less than computing power wasn't as great. But if you wrote enough code, you could finish the job at the end of the day. The sheer volume of data is so significant that writing traditional code-based automation to deal with it has become like an arms race. It's very difficult." - Richard Whitehead